As fixed and mobile computing devices increase in prevalence and processing power, the use of data networks connecting such devices has similarly grown in popularity. However, wired and/or wireless data networks often perform differently under various loading conditions. For example, when served by a heavily loaded data network, a user device can generally experience lower data throughput and higher latency. In some data networks, this network performance degradation can be dependent on the nature of the activity in which the user device is engaged. For example, a user device actively involved in an application, such as requesting and/or viewing a web page or the like, can be more likely to notice data network degradation than a user device that is more passively involved in an application, such as reception of an e-mail or a large file transfer.
Conventionally, data networks generally handle all subscriber devices equally under all circumstances. Accordingly, if network degradation caused by loading and/or other factors nears acceptable limits, such networks often require capacity upgrades to be implemented for all subscribers. However, due to the respective requirements of the subscribers, the subscribers may or may not notice or appreciate these upgrades. As a result, such upgrades can introduce unnecessary costs for a data service provider. Other existing data networks attempt to differentiate between subscribers based on their respective quality of service (QoS) requirements, but the techniques utilized by these networks can be prohibitively complex and/or inefficient and are generally not effective in all cases. Accordingly, it would be desirable to implement techniques for managing a data network that mitigate at least the above shortcomings.